evac error: EVAC line PERSIDNAME problem with IMESH, no mesh found.

[Roger]

hi everyone,

I modified evac_example1a.fds from VTT to run a test case of mine,
then this error popped up and the process was interrupted. So
strange.If there's any need, i can send you the input file.

all the best
Roger

[Roger]

There's a mistake in the title, EVAC line EVACIDname problem with
IMESH, no mesh found.

[Roger]

Is there anyone who met this problem before or can give me some
suggestions?

[enronc]

As I explained to Roger by private e-mail (he sent me the input file),
his problem is that the XB in the EVAC line should belong to the
evacuation mesh.
From the manual: "Defines the rectangle, where the agents are put, z
should belong to the correct main evacuation mesh. If the coordinates
XB intersect many evacuation meshes then the keyword MESH_ID should be
given."

By the way, without having the input file it would have been very hard
to see the problem (for me, maybe Timo could understand it in a
second ;).

Cheers,
Enrico

[Roger]

very appreciate enronchi for helping me fix the problem. Another
problem is that over 10 of all 30 agents were removed from the mesh
while the case was running by warnings being shown with the agents'
xyz coordinates.

[enronc]

Dear Roger,

Have you tried to play with the parameter DENS_INIT? If the problem is
related to too high densities, a higher value of DENS_INIT can help to
put the agents more densily.

Cheers,
Enrico

[TimoK]

Well, I'm back at the office, I was on a skiing holiday.
It is hard to tell anything without the input file, but
as Enrico wrote, the error message is due to the XB. The
EVAC line XB should be "inside" a main evacuation mesh.
The error message could be better, the error messages
are now mainly for me, not for the users. Usually these
errors are due to some mistake in the input file.

And the second problem that you get agents taken away from the
calculation. This is usually due to the problems with OBSTs.
Is you building "sealed" in the evacuation mesh(es)? The
place, where you put your agents using EVAC lines (or ENTR or
moving to different places with DOORs) should have walls
everywhere. You probably have some "holes" in your geometry
so that the agents can go to the outer boundary of the mesh and
when they move away from the mesh they are just deleted and a
warning message is printed. See the agents in Smokeview if this
happens.

Or then you might have error in your definition of DOOR and the
corresponding
"outflow" VENT. If you do not have a DOOR (or an EXIT) in front of an
"outflow" vent, then the agents just move through the "outflow" vent.
The agents do not feel any wall forces at an "outflow" vent boundary.
This way the agents might get in places where they should not be and
you get the error message.

TimoK

[Roger]

My room,i have only one room, is with 2 doors and several windows
only. My evac mesh boundary is right on the outer surface of my
room,and evac doors is made by HOLE commands. And i've sent you the
input file by email. Probably you can tell me where and how i should
modify the lines. Thank you very much indeed !

All the best

Roger

[TimoK]

Ok, I found your email with the input file. I had quite a few
emails in my inbox...

Timo

[TimoK]

Roger, I found the mistake. You had "0.2" and it should
be "10.2", so it was a small typo.

VENT XB= 0.2,11.2, 0.05,0.05, 0.4,1.6, SURF_ID='OUTFLOW',
MESH_ID='MAINEVACGRID', EVACUATION=.TRUE., COLOR='RED' / #
FRONT_DOOR
&VENT XB= 10.2,11.2, 0.05,0.05, 0.4,1.6, SURF_ID='OUTFLOW',
MESH_ID='MAINEVACGRID', EVACUATION=.TRUE., COLOR='RED' / #
FRONT_DOOR

Your vent for the frontexit had wrong XB for the main
evacuation mesh.

Added also "EVACUATION_DRILL=.TRUE." to the MISC line, it
is easier to debug the evacuation input this way. And you
should use it when you are building up your evacuation
input. You see better the actual evacuation geometry,
when you do not have fire meshes present.

&MISC SURF_DEFAULT='WALL', EVACUATION_DRILL=.TRUE.,
EVAC_SURF_DEFAULT = 'EVAC_WALL' /

TimoK

[Roger]

Dear Timok,

Thank you very much indeed for help. I'll test it tomorrow,and let you
know if it's a success.
According to the "EVACUATION_DRILL=.TRUE.", will this debug function
command line auto adjust evacuation geometry and convert it to a more
reasonable one?

[enronc]

Dear Roger,

EVACUATION_DRILL=.TRUE. is not a debug function but it just discards
all fire information and all fire meshes.

Cheers,
Enrico

[Roger]

Now i know it,it's just 'evacuation drill mode', it's been finished
very soon,so i can see any messages of incorrect inputs. Now my input
file doesn't have any problem,thank you all!
Anyway, i found a phenomena. When i play the smv file,the evacuation
route of people to mid door pass the fire,people run through the fire
getting out from the mid door with no dead.Background; the fire
location is close to the mid door. What's the principle cause dead in
fds+evac,CO CO2 O2 mass fraction or directly by fire?

[enronc]

Roger,

FDS+Evac uses the FED concept introduced by Prof. Purser. Please, read
the the manual for having more information about it and also the Prof.
Purser's SFPE chapter.

From the manual: "The present version of FDS+Evac uses only the
concentrations of the narcotic gases CO, CO2, and O2 to calculate the
FED value.. ..An agent is considered to be incapacitated when the FED
value exceeds unity. An incapacitated agent is modelled as an agent,
which does not experience any social forces from the other agents and
walls and whose target movement speed, v0i , is set to zero. The size
of an incapacitated agent is not changed, i.e., it remains on its
feet."

Regarding fire influence, maybe this old post could be interesting for
you:
http://groups.google.com/group/fds-smv/browse_thread/thread/649a8dadcb0749b4/2b35037410e20c37?lnk=gst&q=FED#2b35037410e20c37

Cheers,
Enrico

[Roger]

I've read many papers and technical references recently,have too much
information in my mind. Thank you for reminding me this. I have to
read the manual book again.

[TimoK]

If you have a fire close to the mid door, then the agents
should not choose this door, if you are using the door
selection algorithm present in FDS+Evac. "using the door
selection algorithm" = you should have an additional
evacuation mesh for each door (DOORs and EXITs) so
that the agents can find a way to each different exit
if they want so. And if you have a fire you could put
there an OBST with EVACUATION=.TRUE. so that the agents
are not going to walk over it. Note that the smoke
information is taken at the z-level 1.6 m by default.

TimoK